Application of KLIMA/G-POD algorithm to CO2 retrieval from IASI/METOP-A observations and comparison with GOSAT/TANSO-FTS products

The ESA research project “Application of KLIMA algorithm to CO₂ retrieval from IASI/METOP-A observations and comparison with GOSAT/TANSO-FTS products” aims at the develop of a dedicated software, based on the KLIMA inversion algorithm (originally proposed by IFAC-CNR for the ESA Earth Explorer Mission), optimally suited for CO₂ retrieval and integrated into the ESA GRID-based operational environment G-POD (Grid Processing On-Demand) to processing Level-1 data acquired by the IASI instrument onboard the METOP-A satellite and to perform a comparison and cross-validation of GOSAT TANSO-FTS Level-2 data. Optimized versions of the KLIMA-IASI code have been investigated, aiming at developing a non-operation retrieval code with capabilities that meet the requirements of cross validation with GOSAT TANSO-FTS products and with adequate features for the integration on the G-POD system. The KLIMA algorithm consists of two distinct modules: the Forward Model and the Retrieval Model, optimized for inverse processing of IASI measurements. The Forward Model is a line-by-line radiative transfer model, with capability to simulate wideband spectral radiances acquired by IASI. The adopted spectroscopic database is HITRAN 2008. Dedicated spectroscopic database and line shape are implemented for CO₂, to take into account the line-mixing effect (Niro et al., 2005). The Retrieval Model uses a constrained Nonlinear Least Square Fit approach and the cost function to be minimized takes into account the a priori information (Optimal Estimation Method) and the Marquardt parameter. To reduce the impact of systematic uncertainties, this module implements a multi-target retrieval (more than one species is simultaneously retrieved) and a complete variance-covariance matrix (VCM), including both the measurement errors and the errors in the estimate of FM parameters. For the performance of the retrieval a target accuracy of 0.3% (1 ppmv out of 370 pp- v) on regional scales (1000 × 1000 km) at monthly intervals, which is consistent with the requirements of the GOSAT mission for CO₂ products, was assumed as reference value. The required maximum program size was set to one Gbyte and the running time was limited with the aim of processing one orbit of IASI data in one day when using G-POD computing resources. The KLIMA-IASI ARM algorithm has been successfully integrated in the ESA G-POD for Earth Observation Application and now the code is available to all interested users for bulk processing of IASI data. ESA G-POD is a GRID based-operational environment, accessible using a dedicated Web portal (http://gpod.eo.esa.int/). It hosts more than 350 local computing nodes, more than 500 TB of EO data coming from ESA and non-ESA satellites. The full capacity of G-POD computing resources, that can be made available for KLIMA-IASI bulk processing, are currently under evaluation; moreover the upload of the full set of IASI data acquired in one year (march 2010 - march 2011) shall be completed (3 months are available at the moment). After the procurement of a consolidated version of the KLIMA-IASI/G- POD retrieval code it was possible to start the processing of IASI spectra and for comparison and cross-validation of KLIMA-IASI CO₂ products with GOSAT/TANSO-FTS operational products. Using the KLIMA inversion code integrated into the ESA G-POD, it was possible to perform an extensive inter-comparison and cross-validation of a selected set of IASI measurements collocated with GOSAT TANSO-FTS observations. The KLIMA-IASI code was used for the retrieval of CO₂ total column from a series of test cases for comparison with GOSAT TANSO-FTS SWIR L2 data. Retrieved values from measurements acquired by the two instruments in February 2011 in two selected areas between 25-45°N over Nord America and Asia are shown in figures 1-2, along with maps of the footprints of IASI (red crosses) and GOS